By William K. Black
Quito: June 25, 2015
The BBC came up with a good “hook” for a story on the troika’s assault on the Greek economy and people. “Yanis Varoufakis, the Greek finance minister, spent his academic career … studying game theory.” Professor Marcus Miller, a UK economist (U. Warwick) wrote an article for the BBC premised on how Varoufakis would apply game theory to Greece’s negotiations with the troika (the IMF, ECB, and the European Commission). Miller is a colleague of the great Robert Skidelsky and has co-authored with him an article explaining the economic illiteracy and self-destructive nature of the troika’s (and UK’s) infliction of austerity in response to the Great Recession.
The BBC, however, is such a great fan of austerity that one rarely reads why the vast majority of economists think that using austerity to respond to a Great Recession is akin to the quackery of bleeding a patient to make him healthier. Miller’s article in the BBC about game theory has the wrong title (recall that the author often does not get to choose the title), the wrong game, the wrong concept, and the wrong payoffs. The title of the article is: “Can game theory explain the Greek debt crisis?” The article does address that issue. It is limited to the issue of the new Greek government’s negotiations with the troika concerning a crisis that they inherited.
The game that Miller uses is the “prisoner’s dilemma.” That is the wrong concept and the wrong game and should actually be called the “prisoners’ dilemma” because it requires at least two prisoners. The “prisoners’ dilemma” game is used to explain (1) why cooperative behavior – by criminals – would be their optimal strategy, (2) why prosecutors and the police should prevent that cooperation, and (3) how prosecutors and the police can shape the prisoners’ incentives to encourage them to confess. As conventionally pictured, and Miller falls into this trap, the game does a poor job of explaining the third point. Real life prosecutors, police, and criminologists in the U.S. do a far better job of optimizing the incentives than do economists – and did so long before game theory was developed.
Here is Miller’s explanation.
The most famous game of all is the Prisoner’s Dilemma. Imagine two prisoners have to choose between confessing and staying silent. If they both stay silent, they both go to jail for one year. If one confesses and the other stays silent, the first goes free and the second gets 20 years. If both confess, they both get five years.
As anyone with even the faintest understanding of the U.S. criminal justice system knows, which includes anyone who has watched a U.S. police drama on television, Miller’s description is deeply suboptimal. It misses the key element of timing. The actual dynamic, which optimize the incentive to confess, is whoever rats out his confederate first gets a far better deal. It would be nuts to give both the same sentence (“five years”) if they “both confess.” It is common for both defendants to confess.
But the deeper problem is that this is not the proper game for analysis of either the Greek crisis or Varoufakis’ negotiations with the troika. Greece, and the Greek people, are not criminals, no one is trying to get a confession, and the “game” everyone should be using is a “cooperative” game rather than prisoners’ dilemma – which is premised on preventing cooperation.
If anyone is using prisoners’ dilemma as its game theory in the Greek context it would have to be the troika. The logic would go like this. Greece’s position in favor of a “troubled debt restructuring” (TDR) and an end to austerity is in the interests of the peoples of Europe (and the world). The appropriate cooperative “game” would be to (a) strike a deal to end austerity in response to recessions throughout the EU and (b) to negotiate TDRs in Spain, Italy, and Greece. That cooperative “game,” however, is being blocked by the troika and is bitterly opposed by conservative leaders in Spain, Italy, the Baltic States, and Ireland. Part of this refusal to enter into a win-win cooperative game on the part of the troika is ideological. Part of the troika’s refusal is reputational. If the leaders of the troika were to end austerity and the EU’s economy were to improve more rapidly it would be undeniable that their economically illiterate infliction of austerity had caused massive, gratuitous suffering.
Similarly, the opposition of conservative political leaders in various EU states to employing a cooperative “game” in Greece is that it would threaten their power and reputation. Spain’s conservative party would fall from power in 24 hours if its leaders ever admitted that austerity forced Spain into Great Depression levels of unemployment and that austerity was the problem rather than the solution. The Prime Minister of Spain is eager to have Greece fail to convince the troika to adopt a cooperative “game.” He would love to see the troika force the Greek government to inflict even greater austerity on the Greek people in order to discredit the Greek government and his Spanish anti-austerity opponents, the surging Podemos party.
Miller and Skidelsky know these facts and have put them in writing. If Greece were a corporate debtor, its creditors would have routinely negotiated a TDR with it seven years ago – precisely because normal creditors understand that their interaction with a debtor who cannot repay the full debt on its original terms should take the form of a cooperative “game.” Normal creditors, if the troika is not involved and thinking in extortionate terms, realize that they should cut their losses through a deal that reduces and stretches out the payments and lowers the interest rate on the debt.
This is why a firm like Elliot Management, Paul Singer’s vulture hedge fund, which tries to profit from, and therefore discourage, TDRs – see its attempt to extort Argentina – is so odious and bad for the world. Richard Zabel, the number two (non) prosecutor of the elite banksters that led the three great epidemics of accounting control fraud that drove the crisis in the U.S. Attorney’s office for Manhattan, just announced that he will soon take a job with Singer. In a grotesque act, while still a purported prosecutor, he issued an ode to Singer’s vulture ethics that was quoted by the New York Times. Zabel’s ode is a further proof of our family rule that it is impossible to compete with unintentional self-parody.
“Elliott has a long record of success and relies on the rule of law as a pillar of its investment philosophy,” Mr. Zabel, who will leave his prosecutorial job at the end of June, said in an interview.
Hedge fund owners are the wealthiest people on Wall Street, so this is the new nature of the revolving door that helps produce immunity for elite banksters and the ethics-free world of the hedge funds. The NYT article explained the (non) prosecutors’ quest for “Wall Street riches.”
[Zabel’s] move also reflects a broader shift on Wall Street, where hedge funds and private equity firms, rather than law firms, are increasingly recruiting federal prosecutors, enticing them with the promise of Wall Street riches. Steven A. Cohen’s Point72 Asset Management — the successor to his SAC Capital Advisors, which pleaded guilty to insider trading charges in an investigation overseen by Mr. Zabel — recently hired a former United States attorney for Connecticut.
Zabel was an important contributor to eliminating the “rule of law” for elite banksters and will enjoy “Wall Street riches” precisely because Singer views him as helping to fend off any (unlikely) effort by the government to rein his depredations. In game theoretic terms, Singer’s strategy is designed in a manner that discourages cooperative “games” that make the world better off.
The oddest aspect of Miller’s description of his hypothetical application of what he terms the “prisoner’s dilemma” to Greece’s negotiations with the troika are the “payoffs” he assumes to the different decisions. His description makes no sense given his article with Skidelsky.
Imagine Greece moves first to avoid default by putting a plan on the table. This plan involves new taxes on the wealthy and changes to pensions – avoiding spending cuts and and having some of its debts written off in exchange. If this plan is accepted by the rest of the eurozone, then Greece is content. Let’s give its payoff a score of 1.
To work out how the rest of the eurozone will respond, one has to see what they stand to gain by accepting Greece’s plan, or by rejecting it.
If the eurozone accepted this deal, the monetary union would remain intact, but it would have to ease its strict rules on fiscal policy and take a loss on holdings of Greek debt. Let’s give the eurozone payoff a score of ¾. So the overall payoff is (1, ¾ ).
But Miller and Skidelsky have explained why these payoffs are incorrect. They have explained that if the EU were to end austerity as a response to a Great Recession the results for the EU would be enormously positive. So, the payoff from a cooperative “game” with Greece that would end self-destructive austerity would be exceptionally positive for the EU and positive for Greece. The EU is much larger than Greece, so its payoff from cooperating with Greece would be far larger than Greece’s payoff. Miller also knows that it is wrong to think that the EU would lose due to “having some of its debt written off in exchange.” Like other creditors, the EU would get more from cooperating with Greece and negotiating a TDR than forcing Greece into a default on its debts. TDRs are negotiated routinely in the private sector because cooperative “games” benefit both parties. So, “the overall payoff” of a cooperative “game” with Greece would be something on the order of (1, 5). Both Greece and the EU would gain, but the EU’s gain would be much larger. The EU’s gain would be greater still if it were to enter into similar cooperative “games” with Spain, Italy, Portugal, and the Baltic states.
Austerity is a coercive, non-cooperative “game” that makes both parties worse off (but aids the EU’s worst politicians). It is terrible economics, and represents a “dismal [non] science.” Game theory is largely devoted to demonstrating the benefits of cooperation. Precisely because Varoufakis is an expert in game theory, and as we can see throughout the new Greek government’s negotiations with the troika, the effort has been to try to get the troika to escape from its self-destructive dogmas and debilitating political urge to punish the Greek people, and see that it is in the interests of the Greek people and the peoples of the EU to cooperate by ending austerity and negotiating a TDR.
The troika’s game theoretic strategy in dealing with the 100 million people of Spain, Italy, and Greece that it has gratuitously forced into Great Depression levels of unemployment is a nasty variant of the “Dictator” game. The conventional dictator game works like this. The dictator makes an offer to divide up $1. Economists define his “rational” offer as 99 cents for him, and one penny to the other player. The other player gets only one choice – he can refuse or accept the dictator’s offer. If he rejects the offer he gets nothing. Economists define the recipient’s rational response as accepting the penny. But there are three striking results of studies of people’s response to the dictator game. First, when the dictator offers a penny, or any very unfair division, people overwhelmingly reject the offer even though they know that this will mean they get nothing. Second, people playing the role of the dictator typically offer to split the $1 evenly, which leads to routine acceptance. Third, the exception to this result is economists and economics majors. They are much more likely than human beings to respond by being nasty little dictators and passively accepting drones who agree to take a penny. Normal human beings care a great deal about fairness and are willing to suffer personal losses rather than give in to dictators like the troika.
The language of the troika and of German politicians about the troika’s mode of dealing with the Greek government is strikingly similar to the dictator game. But there is a vital way in which the troika’s game theoretic approach is far worse than the dictator’s game. In the conventional dictator’s game the other party is made better off, albeit by only a penny. The troika insists on making Greece worse off by insisting on austerity and trying to block a cooperative TDR that would benefit the creditors and the debtors. The game that the troika is playing against the people of Grace and the EU embraces all the imperialism inherent in the dictator’s game – but it is a “negative sum” game that makes the peoples of Greece and the EU worse off. Cooperative games typically are “positive sum” games that make both parties better off.